Update app.py

app.py

@@ -40,14 +40,14 @@ spaBERT_model.eval()
 #Load data using SpatialDataset
 spatialDataset = PbfMapDataset(data_file_path = data_file_path,
                                         tokenizer = bert_tokenizer,
-                                        max_token_len = 256,              #Originally 300
+                                        max_token_len = 256,                          #Originally 300
                                         #max_token_len = max_seq_length,              #Originally 300
                                         distance_norm_factor = 0.0001,
                                         spatial_dist_fill = 20,
                                         with_type = False,
                                         sep_between_neighbors = True,    
                                         label_encoder = None,             
-                                        mode = None)                      #If set to None it will use the full dataset for mlm
+                                        mode = None)                                  #If set to None it will use the full dataset for mlm
 
 data_loader = DataLoader(spatialDataset, batch_size=1, num_workers=0, shuffle=False, pin_memory=False, drop_last=False) #issue needs to be fixed with num_workers not stopping after finished
 
@@ -81,7 +81,7 @@ def process_entity(batch, model, device):
     #pivot_embeddings = embeddings[:, :pivot_token_len, :]
 
     #return pivot_embeddings.cpu().numpy(), input_ids.cpu().numpy()
-    return embedding.cpu().numpy(), input_ids.cpu().numpy()
+    return embedding, input_ids
 
 all_embeddings = []
 for i, batch in enumerate(data_loader):
@@ -112,6 +112,12 @@ def get_bert_embedding(review_text):
 
 
 
+
+
+
+
+
+
 st.title("SpaGAN Demo")
 st.write("Enter a text, and the system will highlight the geo-entities within it.")
 
@@ -153,9 +159,11 @@ if st.button("Highlight Geo-Entities"):
         # Debug: Print the embeddings themselves (optional)
         st.write("Embeddings:", bert_embedding)
 
+        #combine the embeddings (NOTE: come back and update after testing)
         combined_embedding = torch.cat((bert_embedding,all_embeddings[0]),dim=-1)
         st.write("Concatenated Embedding Shape:", concatenated_embedding.shape)
         st.write("Concatenated Embedding:", concatenated_embedding)
+        
         # Process the text using spaCy
         doc = nlp(selected_review)